> ## Documentation Index > Fetch the complete documentation index at: https://apione.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. # Usage, costs, and rate limits > Understand group pricing and token charges, review usage, and manage concurrency. Use the APIONE console to check request usage and charges. The `usage` field in API responses can be used to observe token consumption; actual charges are based on platform usage records and bills. ## Review consumption Open the usage section of the [console](https://apione.ai/en/app). Filter by available time ranges, keys, and models. Use request IDs to locate calls in trends and request details. Fields and refresh timing depend on the current console. ![APIONE usage page request details](https://mintcdn.com/apione/pAFvFGkbmcRZirTZ/images/ui/request-logs.png?fit=max\&auto=format\&n=pAFvFGkbmcRZirTZ\&q=85\&s=093db383e228008dc3c42cb6a7849d28) Request details can be filtered by time, source, API Key, and request ID, and you can view status, latency, tokens, and charges. ## Tokens and charges Chat responses commonly include prompt\_tokens, completion\_tokens, and total\_tokens. Some models expose cache usage. Input, output, cached tokens, and media can have different prices, so multiplying all tokens by one rate may be incorrect. > [!NOTE] > > Use current model details, console information, and applicable terms for prices, currency, payment methods, minimum payments, refunds, and cache discounts. For the points program, see [Points and referral rewards](https://apione.ai/docs/en/account/rewards.md). ## Group pricing and billing rules ### What is group pricing? APIONE can set different prices for the same model across billing groups, supporting individual developers, enterprise customers, and different use cases. Each request uses the model prices associated with the billing group selected for its API Key. ### How groups affect prices | Group type | Typical use and pricing | | - | - | | Default group | Standard pricing for regular model calls. | | Discount group | Reduced prices for selected models, available to eligible users. | | Dedicated group | Custom model pricing for enterprise or partner customers. | Actual group names, available models, and prices are shown on the platform. A group may support only a subset of the platform's models. ### Calculate token-based charges For models billed by tokens, use the following formulas when input and output prices are quoted per million tokens: ```text theme={null} Input charge = Input tokens ÷ 1,000,000 × Input price Output charge = Output tokens ÷ 1,000,000 × Output price ``` The base token charge is the sum of input and output charges. Use the model prices for the billing group selected on the API Key. Cached tokens, additional billable items, and model-specific pricing rules may affect the final charge and must be accounted for separately. ### Calculation example For 10,000 input tokens and 2,000 output tokens, with input and output prices P and Q per million tokens, the base charge is `0.01 × P + 0.002 × Q`. This illustrates the formula; replace P and Q with the actual model and group prices. ### Select a billing group 1. Open API Key management in the [APIONE console](https://apione.ai/en/app/api-keys). 2. Create or edit an API Key and select an available billing group. 3. Review the group's supported models and their input and output prices. 4. Save the configuration and make requests with that API Key. > [!NOTE] > > The same model may have different prices across groups. Group prices can change; check the latest prices shown on the platform. Usage records and bills determine the actual charges. ## Plan credits and pay-as-you-go balance APIONE plans are **one-time purchases, not automatically renewing subscriptions**. The full credit allowance is issued at purchase, not daily. Every plan is valid for **28 days**. | Plan | Purchase price | Credits issued once | Validity | | - | - | - | - | | Plus | \$20 | \$23 | 28 days | | Pro | \$99 | \$120 | 28 days | | Max | \$240 | \$300 | 28 days | The pay-as-you-go balance is separate from plan credits and is unaffected by upgrades or plan expiry. At expiry, the plan ends and unused plan credits are cleared. Buy another plan yourself if you want to continue; no renewal payment is taken automatically. ### Upgrade during the validity period You can upgrade at any time while the plan is valid. Pay the difference between the new and current plan prices, prorated by the time remaining. The upgrade takes effect immediately and keeps the original expiry time. It does not restart a 28-day period. For example, upgrading a Plus plan after 20 days leaves 8 days of validity. Keep the current plan's unused credits and add only the prorated difference between the two plans' total credit allowances. A full new allowance is not issued again. Each subsequent upgrade uses the current tier immediately before that upgrade, preventing repeated credit grants. ```text theme={null} Upgrade charge = (New plan price - Current plan price) × Time remaining ÷ 28 days Additional credits = (New plan total credits - Current plan total credits) × Time remaining ÷ 28 days Available credits after upgrade = Current plan unused credits + Additional credits ``` If Plus has 8 days left and \$3 of unused plan credits, upgrading to Pro costs approximately \$22.57, adds approximately \$27.71 of credits, and leaves approximately \$30.71 available until the original expiry time. The example uses whole days and rounds to cents; the upgrade confirmation shows the actual remaining-time calculation and settlement precision. | Item | Calculation result | | - | - | | Upgrade charge | (\$99 - \$20) × 8 ÷ 28 ≈ \$22.57 | | Additional credits | (\$120 - \$23) × 8 ÷ 28 ≈ \$27.71 | | Available credits after upgrade | \$3 + \$27.71 ≈ \$30.71 | | Remaining validity | 8 days, keeping the original expiry time | ### Downgrades and expiry Downgrades are not supported. Wait for the current plan to expire, then select and purchase another plan. Only unused plan credits are cleared at expiry. The separate pay-as-you-go balance remains unaffected, and the plan does not renew automatically. ### Recharge verification 1. Select an amount on the recharge page and enter the checkout. Verify the actual payment currency, amount, and payment method. 2. After payment, return to APIONE and check the order status and balance change. 3. If the order is still processing, query the original order and keep the payment receipt. 4. When support is needed, provide the order number, payment time, and amount. Avoid submitting the full payment account or API Key. Points redemption and referral rewards follow the [reward rules](https://apione.ai/docs/en/account/rewards.md). Points validity and plan validity are calculated separately. ## Control concurrency APIONE limits concurrent requests per API Key. Upstream rate limits can also apply. Use the concurrency quota configured for your account. Use a queue or semaphore for batch work. For 429 responses, honor Retry-After when present and use bounded exponential backoff with jitter. Avoid infinite retries or creating extra keys to bypass limits. Do not silently restart interrupted streams, which can duplicate generation and charges. ## Failed, cancelled, and interrupted calls Model calls finally classified as failed by the platform are not charged. A client timeout, cancellation, disconnected stream, or failed download after successful generation does not by itself mean the platform call failed. If generation completed or billable usage was confirmed, the applicable model and group rules determine the charge. For streaming, check the final status and actual usage rather than HTTP 200 alone, the client's error message, or whether the client received every output chunk. Incomplete delivery does not prove that no generation occurred. For a disputed status or charge, retain the request ID, timestamp and timezone, model, final status, and error details, then contact support. Confirmed erroneous or duplicate charges are corrected after investigation. Avoid repeatedly creating tasks while the original outcome is unknown. | Situation | Classification and billing | | - | - | | Authentication, validation, or rate-limit checks reject a request before model execution | No model-call charge is incurred for that request | | The platform finally confirms a failed model call | The failed model call is not charged | | The client cancels, times out, or disconnects after receiving part of a stream | This alone does not establish a platform failure; check final status and confirmed usage | | Image or video generation succeeds but the download fails | Generation and delivery are separate; a download error does not automatically reverse the completed generation charge | | A video task exists but the final outcome is unknown | Query the original task and retain its ID rather than creating repeated tasks after client timeouts | ## Cost checklist * Limit unnecessary conversation history and maximum output. * Compare quality and usage on a small representative sample first. * Separate testing and production keys and investigate unusual activity. * For client errors, check the platform's final status before retrying. Avoid duplicate requests while the outcome is unknown. References: [Usage](https://apione.ai/en/api-docs/docs/usage) and [rate limits](https://apione.ai/en/api-docs/docs/rate-limits). This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.